Albacore

🚀 MCP Foundation - Universal Model Context Protocol Framework

This folder provides a generic implementation foundation for building the Model Context Protocol (MCP) in AI-driven applications. It offers a standardized approach to creating MCP servers and clients, enabling the integration of LLMs into your applications.

✨ Features

• Standardized MCP Server: A basic server implementation supporting HTTP and stdio transport methods.
• Universal MCP Client: A client for connecting to any MCP server.
• Ollama Integration: A ready-to-use service for generating embeddings and text (via Ollama).
• Supabase Integration: Built-in support for the Supabase vector database.
• Modular Design: A clear organizational structure for resources, tools, and prompts.
• Example Templates: Example implementations to help you get started quickly.

📁 Directory Structure

_mcp-base/
├── server.ts            # Main MCP server implementation
├── client.ts            # Universal MCP client
├── utils/               # Utility services
│   ├── ollama_embedding.ts    # Embedding generation using Ollama
│   └── ollama_text_generation.ts  # Text generation using Ollama
├── tools/               # Tool implementations
│   └── sample-tool.ts   # Example tool template
├── resources/           # Resource implementations
│   └── sample-resource.ts  # Example resource template
├── prompts/             # Prompt implementations
│   └── sample-prompt.ts # Example prompt template
└── README.md            # This document

🚀 Quick Start

Prerequisites

• Node.js and npm/pnpm
• Ollama for local embedding and text generation
• A Supabase account for vector storage

Environment Configuration

Create a .env file and add the following variables:

PORT=3000
SUPABASE_URL=https://your-project.supabase.co
SUPABASE_SERVICE_KEY=your-service-key
OLLAMA_URL=http://localhost:11434
OLLAMA_EMBED_MODEL=nomic-embed-text
OLLAMA_LLM_MODEL=llama3
SERVER_MODE=http  # 'http' or 'stdio'

Server Initialization

1. Import the required modules.
2. Register your resources, tools, and prompts.
3. Start the server.

// Import the base server and utilities
import server from "./server";
import { registerSampleResources } from "./resources/sample-resource";
import { registerSampleTool } from "./tools/sample-tool";
import { registerSamplePrompts } from "./prompts/sample-prompt";

// Initialize the database (if necessary)
async function initializeDatabase() {
  // Your database initialization logic
}

// Register your components
registerSampleResources(server, supabase);
registerSampleTool(server, textGenerator, embeddings, supabase);
registerSamplePrompts(server, supabase);

// Start the server
startServer();

Client Usage

import MCPClient from "./client";

// Create a client instance
const client = new MCPClient({
  serverUrl: "http://localhost:3000",
});

// Example of invoking a tool
await client.invokeTool("sample-tool", { input: "test" });

// Example of reading a resource
const resourceData = await client.getResource("sample-resource");

// Example of getting a prompt
const prompt = await client.getPrompts();

🛠️ Extending the Framework

To add new features, you can follow these steps:

1. Create a New Tool: Create a new tool file in the tools/ folder and export the tool function.
2. Register the Tool: Import and register your new tool in server.ts.
3. Use the Tool: Clients can use your new tool by calling the invokeTool method.

📄 License

This project is licensed under the MIT License.